From: Mika Westerberg Subject: Re: [PATCH v6] pci: prevent putting nvidia GPUs into lower device states on certain intel bridges Date: Fri, 6 Mar 2020 07:40:06 +0200 Message-ID: <20200306054006.GZ2540@lahna.fi.intel.com> References: <20200303101052.133631-1-kherbst@redhat.com> <20200304093324.GI2540@lahna.fi.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-pci-owner@vger.kernel.org To: Karol Herbst Cc: LKML , Bjorn Helgaas , Lyude Paul , "Rafael J . Wysocki" , Linux PCI , Linux PM , dri-devel , nouveau List-Id: nouveau.vger.kernel.org On Thu, Mar 05, 2020 at 05:11:57PM +0100, Karol Herbst wrote: > On Wed, Mar 4, 2020 at 10:33 AM Mika Westerberg > wrote: > > > > Hi, > > > > On Tue, Mar 03, 2020 at 11:10:52AM +0100, Karol Herbst wrote: > > > Fixes state transitions of Nvidia Pascal GPUs from D3cold into higher device > > > states. > > > > I think it is good to explain bit more here why this fix is needed. > > > > is something like this fine? > > Fixes the infamous 'runpm' bug many users are facing on Laptops with Nvidia > Pascal GPUs by skipping PCI power state changes on the GPU. I would open up 'runpm' -> runtime PM. > It's still unknown why this issue exists, but this is a reliable workaround > and solves a very annoying issue for user having to choose between a > crashing kernel or higher power consumption of their Laptops. Also I think it would be good to include a short log snippet how this manifests and in what kind of scenario it happens.